Composite score
–/100
Breakdown
Method
The score is the weighted mean of five components, each normalized 0–100.
If a data source is unavailable for a cycle, its component is dropped and the
remaining weights are renormalized.
- Staking ratio · weight 20%
- Bonded voting power of the consensus set ÷ circulating supply, compared to the
Namada target staking ratio of 40%. Score =
min(100, ratio / 0.40 × 100). Circulating supply is read from the indexer at the latest completed epoch. The numerator counts consensus-set stake only (it excludes stake bonded to below-capacity validators and tokens in unbonding), so it slightly understates the true staking ratio. - Nakamoto coefficient (33.4%) · weight 25%
- Minimum number of top validators whose combined voting power reaches the
consensus liveness threshold (≥ 1/3 offline). Higher is better.
Bands:
NC ≥ 8 → 100,6–7 → 85,5 → 70,4 → 55,3 → 40,2 → 20,1 → 0. - VP concentration (HHI) · weight 15%
- Herfindahl–Hirschman index of voting-power shares (sum of squared shares,
on a 0–10000 scale). Score maps linearly from
HHI 500 → 100(roughly a uniform 20-validator set) toHHI 2500 → 0. - Liveness margin · weight 25%
- From the latest canonical commit sample: signing voting power minus the 66.7%
required to keep finalizing blocks. Score =
clamp(margin / 15% × 100, 0, 100); a margin ≤ 0 means consensus would be paused. - Offline scenarios (top set) · weight 15%
- Counts of top-validator scenarios that would pause consensus:
Score =
max(0, 100 − 15 × #single-validator critical (top 10) − 5 × #pair critical (top 8)).
Score bands
| Score | Band |
|---|---|
| 80 – 100 | Healthy |
| 60 – 79 | Good |
| 40 – 59 | Degraded |
| 0 – 39 | Critical |